Dr. Leyenaar is a pediatric hospitalist and Assistant Professor of Pediatrics at Tufts University School of Medicine. Her goal is to become an independent clinician-investigator and a national leader in the care of hospitalized children, focused on improving health care efficiency while providing patient-centered care. Dr. Leyenaar's proposed career development plan incorporates mentored research and training in advanced statistics, qualitative methods, stakeholder engagement and implementation research as part of a PhD in Clinical and Translational Science. Dr. Leyenaar's mentorship team brings expertise in patient-centered outcomes research in a comparative effectiveness framework, guiding her career development. Dr. Lindenauer, primary mentor, is an accomplished investigator with an established record of mentoring investigators toward independent research careers. Dr. Leyenaar works in an environment extremely supportive of her success, with access to all of the resources required to achieve her long-term goals. In 2012, 1.8 million children were hospitalized in the United States, generating aggregate costs of 20.6 billion dollars. The majority of these children were admitted to hospitals through emergency departments (ED). For children who are medically stable, direct admission is an alternative approach involving referral to hospital without receiving care in the hospital's ED. Among children with timely access to outpatient healthcare providers, direct admission may offer benefits including improved coordination between health care providers, reduced ED volumes, and reduced healthcare costs. However, potential risks of direct admissions include delays in initial evaluation and management that might adversely impact patient safety. The goal of Dr. Leyenaar's research is to identify the pediatric populations, healthcare settings, and procedures that facilitate safe and effective direct admissions. She will accomplish this goal by: (1) comparing the outcomes of direct and ED admissions for common clinical conditions; (2a) characterizing the perspectives of stakeholders in the direct admission process; (2b) developing a consensus-driven guideline and identifying safety and quality indicators for direct admissions; and (3) pilot test the feasibility and acceptability of implementing this direct admission guidelin. Aim 1 will be accomplished using a large administrative database to conduct a retrospective cohort study of children hospitalized at structurally diverse hospitals in the United States, developing advanced statistical models to compare outcomes among children admitted directly to those admitted via EDs. Aim 2 will be accomplished by engaging parents, healthcare providers, payers and policymakers in deliberative discussions about their admission experiences and perspectives, followed by a Delphi process to develop a direct admission guideline and prioritize quality indicators. Aim 3 will involve implementation of this guideline at one hospital for six months, generating data vital to subsequent studies and Dr. Leyenaar's emergence as an independent investigator.